Behind the Scenes: A Technical Deep Dive – Chasing Innovation From the Ground Up

What was the challenge, and how did the geological conditions contribute to it?

The Méquillon Deposit is a magmatic Ni-Cu-PGM deposit, related to a suite of mafic to ultramafic intrusions that occur throughout the Nunavik Nickel Project in Quebec. Méquillon consists of disseminated and net-textured sulfides at the base of a composite, steeply dipping ultramafic intrusive body. The body intruded in close proximity to strongly conductive pyrrhotite- and graphite-rich argillites within the sedimentary country rock. The combination of disseminated sulfide mineralization, strongly resistive host intrusion, and highly conductive layers within the country rock obscure the EM signature of the deposit. 

The challenging nature of the mineralization, specifically the lack of a massive sulphide core, is why Canadian Royalties decided to test new exploration tools — in this case, a ZTEM survey. 

We were mandated to perform a 3D inversion of the ZTEM data, to validate the ore body had a clear signature in the inversion model. The challenge was to recover the deposit signature where adjacent to a highly conductive zone.

Why didn’t the UBC code work? 

The UBC-GIF MTZ3D program was first used to perform the inversion. This code uses a rectilinear mesh for natural sources. Due to limitations related to the problem size, it led to poor data fit of Méquillon’s signal seen in the data. We decided to test the open-source SimPEG program, to see if the results could be improved.

What is the SimPEG code, and why was it effective for this case study?

The SimPEG python package is built for generic forward modelling and inversion of geophysical data, but in this case was used for the 3D inversion of ZTEM data. The inversion code uses a direct solver rather than an iterative method to solve the forward simulation. In turn, this provides greater accuracy of the forward problem, which can propagate into the inversion results.

The program also requires use of an ocTree mesh – a type of adaptive meshing that can help resolve numeric problems with efficient discretization around survey stations and in regions of high topographic relief. It also allows for an adequate cell size when dealing with a wide range of frequencies and anomalies. 

What were the results? 

Geoscience ANALYST Pro Geophysics was used to create the ocTree mesh and input files to run the SimPEG-based inversion. It also allowed us to submit the inversion and monitor its progress on the Azure Cloud. 

The SimPEG inversion results greatly improved the data fit, particularly of the subdued response characterizing Méquillon. Using this geophysical inversion library allowed us to recover the signature of the mineralization at surface and at depth — while mapping in further detail the formational conductors, compared to the model produced by the MTZ3D code. Geoscience ANALYST Pro Geophysics allowed us to easily prepare and run the inversion files, and view and analyze the inversion results.